JSON-LD in 2026: Which Schema.org Types Actually Matter and Which Are a Waste of Time

Schema.org defines over 800 types. Eight hundred. There are types for Volcano, MedicalCondition, MusicComposition, and Cemetery. You could spend months marking up every entity on your site with meticulous JSON-LD — and Google would ignore 90% of it. I know because I have been there. I once spent a weekend adding CreativeWork, WebPage, and CollectionPage schema to every template on a client's site. The result? Absolutely nothing changed in search. No new rich results, no ranking improvement, no visible difference whatsoever. That weekend was wasted.

The uncomfortable truth about structured data is that most of it does nothing. Google only renders rich results for a specific, documented list of types. Everything else gets parsed, maybe stored internally, and never surfaces to users. The question every developer should ask before adding JSON-LD is not "does Schema.org support this type?" but "will Google actually do something with it?"

---

The Complete List of Google-Supported Rich Result Types

As of 2026, Google supports rich results for the following Schema.org types. This list comes directly from Google's documentation and is the only list that matters for SEO purposes.

Schema.org Type	Rich Result	Who Should Use It
Article / NewsArticle / BlogPosting	Enhanced listing with headline, date, author	Blogs, news sites, content publishers
BreadcrumbList	Structured path instead of URL	Every site with navigation hierarchy
Event	Date, venue, ticket info in search	Event organisers, venues, ticketing platforms
FAQPage	Expandable Q&A accordion	Any page with genuine FAQ content
HowTo	Step-by-step instructions inline	Tutorials, guides, DIY content
JobPosting	Listing in Google for Jobs	Job boards, company career pages
LocalBusiness	Google Maps, business card in search	Physical businesses with a location
Product + Review	Price, rating, availability, reviews	E-commerce, product review sites
Recipe	Rich recipe card with image, time, rating	Food blogs, recipe sites
SoftwareApplication	App rating and price in search	App developers, software companies
VideoObject	Video carousel, key moments, thumbnails	Sites hosting video content
WebSite + SearchAction	Sitelinks search box	Established sites with internal search
Organization	Logo, contact info in knowledge panel	Companies, brands
Course	Course listing with provider, price	Educational platforms
Dataset	Dataset search integration	Research, data portals
Book	Book info in search	Publishers, bookstores
MusicGroup / MusicAlbum	Music knowledge panel	Artists, labels, music platforms
Movie	Movie details in search	Movie sites, reviews, cinemas
ProfilePage	Author profile enrichment	Sites with author pages
DiscussionForumPosting	Forum post in search	Forums, Q&A sites

Everything not on this list produces no visible change in Google search. Types like WebPage, CreativeWork, CollectionPage, Service, Person (outside of ProfilePage), Place (standalone), Action, and hundreds of others are technically valid Schema.org markup but have zero effect on how Google displays your pages.

---

The ROI Breakdown: Where to Spend Your Time

Not all supported types deliver equal value. Some dramatically change your search appearance, others provide subtle improvements. Here is how I rank them by return on implementation effort.

Tier 1: Implement Immediately

These types deliver the highest visible impact for the least effort.

Product with AggregateRating and Offer — if you sell anything online, this is non-negotiable. Price, star ratings, and availability badges directly in search results. The CTR difference between a plain listing and a product rich result is 2-3x in my experience. E-commerce sites without Product schema are leaving money on the table every day.

LocalBusiness — for any business with a physical location. Feeds directly into Google Maps, the Google Business knowledge panel, and local search results. If customers visit your location, this is the single most impactful schema type you can add.

BreadcrumbList — trivial to implement, improves every page. Replaces the raw URL in search results with a clean navigation path. Takes five minutes to add to a template and benefits every page on your site. Use our JSON-LD Generator to create the markup.

Tier 2: High Impact for Content Sites

Article / BlogPosting — shows author, date, and headline in enhanced listings. Important for content sites, blogs, and news publishers. The author attribution also builds E-E-A-T signals that Google uses for quality assessment.

FAQPage — generates an expandable accordion in search results, dramatically increasing your visual footprint. Google tightened eligibility in 2023 — FAQ rich results now primarily appear for well-known authoritative sites in general search. But the markup is still consumed by AI systems and may appear in other Google surfaces.

Recipe — if you publish recipes, this is mandatory. Recipe rich results are among the most visually striking in Google search — large image, cooking time, rating, calorie count. Food blogs without Recipe schema are invisible in recipe search.

JobPosting — feeds into Google for Jobs, a dedicated job search experience within Google. If you have a careers page or run a job board, this is essential.

Tier 3: Worth Implementing If Relevant

HowTo — step-by-step rich results for tutorial content. Effective but less commonly triggered than it used to be. Still valuable for technical documentation and DIY guides.

Event — date, venue, and ticket information in search. Important for event organisers and venues. Google also surfaces events in a dedicated events search experience.

VideoObject — video thumbnails and key moments in search. If you host video content (not just embedded YouTube), this helps Google index and display your videos properly.

SoftwareApplication — app rating and price in search. Relevant for software companies and app developers.

WebSite + SearchAction — enables the sitelinks search box (a search bar directly in your Google listing). Only works for established sites that Google already shows sitelinks for. If you do not have sitelinks yet, adding this schema will not create them.

Tier 4: Skip Unless You Are in That Exact Niche

Course, Dataset, Book, Movie, MusicGroup — these only matter if you are in education, data publishing, book retail, cinema, or music. For everyone else, they are irrelevant.

ProfilePage, DiscussionForumPosting — niche types for sites with author profiles or forum content.

---

What About Types Google Does Not Support?

Here is where I disagree with the "just add everything" approach that some SEO guides recommend. Schema.org has types for Volcano, MedicalProcedure, GolfCourse, and LodgingReservation. You could mark up every entity on your site with precise, spec-compliant JSON-LD. But should you?

No. And here is why:

1. No visible result. If Google does not render a rich result for the type, users see exactly the same search listing they would without the markup. The effort produces zero user-facing benefit.

2. Maintenance cost. Structured data is not write-once-forget. Schema changes, pages get updated, templates get refactored. Every JSON-LD block is code that can break, go stale, or produce validation errors. More markup means more maintenance.

3. False sense of completeness. Adding unused schema types gives teams the illusion that they have "done SEO". Meanwhile, the types that actually matter — Product, LocalBusiness, BreadcrumbList — remain unimplemented because nobody prioritised them.

4. Validation noise. Google Search Console reports structured data errors. If you have hundreds of pages with unsupported types and minor validation issues, the real errors in your supported types get buried in noise.

The correct strategy: implement Google-supported types thoroughly and ignore everything else. A site with perfect Product, BreadcrumbList, and Article schema will outperform a site with fifty Schema.org types that are all slightly broken.

---

The One Exception: AI Search Engines

There is a legitimate reason to consider structured data beyond Google's rich result list — and it is a recent one. AI-powered search engines like Perplexity, ChatGPT Browse, and Google AI Overviews consume structured data more broadly than Google's traditional search does.

When an AI system retrieves your page to answer a question, it reads the full HTML including JSON-LD blocks. Structured data helps AI understand the semantic meaning of your content — who wrote it, when, what type of content it is. This can influence whether your page gets cited in an AI-generated answer.

I covered this in detail in the SEO for AI search article, but the practical implication for structured data is:

• Article with datePublished, dateModified, and author helps AI assess freshness and authority
• FAQPage maps directly to how users query AI systems — Q&A pairs are easily extractable
• Organization helps AI attribute content to a known entity

Does this mean you should add every Schema.org type for AI? No. The same ROI logic applies. Focus on the types that both Google and AI systems consume: Article, FAQPage, Organization, Product, BreadcrumbList. These are the types where the implementation effort pays off across multiple channels.

The AI search angle is worth monitoring. If Perplexity and ChatGPT start giving measurable citation preference to pages with specific structured data types, the calculus changes. But as of 2026, the traffic from AI search is still a fraction of traditional Google search for most sites. Invest accordingly.

---

A Prioritisation Framework

When deciding which Schema.org types to implement, run through this checklist:

1. Is the type on Google's supported list? If no, skip it (unless it is Article/Organization for AI reasons).
2. Does the rich result apply to your content? Recipe schema on a blog post about cooking — yes. Recipe schema on a software documentation page — no.
3. Do you have the data to fill required properties? Product schema without price, rating, or availability produces validation errors. Do not add schema for data you do not have.
4. Can you maintain it? If the data changes frequently (prices, availability, event dates), you need automated generation, not manual JSON-LD. If you cannot automate it, the data will go stale and produce errors.
5. What is the expected impact? Use the tier list above. Start with Tier 1, then Tier 2, then decide if Tier 3 is worth the effort for your specific site.

---

The Practical Minimum

For most websites, here is the structured data I recommend:

Every site:

• BreadcrumbList on every page (trivial to automate in your template)
• Organization on the homepage
• WebSite + SearchAction on the homepage (if you have internal search)

Content sites (blogs, documentation):

• Article on every post with datePublished, dateModified, author
• FAQPage on pages with genuine Q&A content

E-commerce:

• Product with Offer and AggregateRating on every product page

Local businesses:

• LocalBusiness with full address, hours, and contact info

Generate all of these with the JSON-LD Generator. Validate with Google's Rich Results Test. Check your meta tags and Open Graph markup at the same time — structured data is one layer of a three-layer metadata stack (HTML meta, Open Graph, JSON-LD) that I described in the structured data implementation guide.

---

The Bottom Line

Schema.org has 800+ types. Google supports roughly 20 for rich results. Your implementation effort should go entirely into those 20. Within those 20, prioritise by tier — Product, LocalBusiness, and BreadcrumbList first, niche types last. Skip everything Google does not support unless you have a specific reason to believe AI search engines will reward it. The goal is not Schema.org completeness — it is search visibility. Focus your structured data where it actually moves the needle.